Saltar al contenido

Etiquetagem morfossintática automática de corpora de língua falada transcrita

We describe in this work the corpus FRACAS-BR, a preliminary translation of the FraCaS corpus, which had the original goal of benchmarking a large selection of English semantic phenomena as far as inference is concerned. We hope to investigate whether these semantic phenomena have corresponding ones in Portuguese and whether native speakers agree with the official inference annotations.